«

»

Jan 28

The System That Trades When I Wouldn’t

Early in 2007 I developed a fairly simple trend following system. It’s a reflection of how I look at the investment world and covers Stocks (S&P, Nasdaq), Bonds (30yr, 10yr), Currencies (Dollar, Yen, Euro, Sterling), and Commodities (Oil, Gold). I’ve maintained signals on it ever since, although I’ve never been able to trade off them as I’d always been regulated and registered as a broker. Pity, because it’s annualized 12.6% since inception.

One of the things about it that has always appealed to me is that it takes signals when I wouldn’t. When I say that I really mean the old me, the sales trader who fancied himself as a fund manager, the prop trader sitting on a desk at Lehman who was instilled with a need to be right, and who didn’t yet understand that some of the most successful fund managers out there only win about 40% of the time, but the size of their winners far outweigh their losers. They’re called trend followers.

Once you’re comfortable with a trend following system that suits your personality and trading style the hard bit is having the discipline to execute it. (That in itself is actually an edge, and a topic for another post.) You already know it’s a system with a positive expectancy but when you’ve watched something climb inexorably and the old you is saying ‘I think we could be close to a top here’, guess what, ping! Here’s a buy signal from your trend-following system. OMG. Seriously? I can’t believe it’s telling me to buy this thing here…

Take crude oil. This is the system that watched oil climb until it had nearly touched $100 for the first time in history and went ping! – long above $99 and got stopped out two weeks later for a small loss. (It later got back in at $91 and rode it to $121). And it’s the same system that watched crude slide from near $150 and just as everyone was calling for a bottom, ping! – short at $82, and covered near the low at $45, nice, crowd wrong, system right.

That’s why I like it, it’s counter-intuitive to my old self, it’s not contrarian in that it’s not picking tops or bottoms, it’s going with a well-established trend just as many consider it may be ending. I’ve seen it happen so many times now it amuses me. I actually welcome the conflict it presents because I know if the system is wrong it won’t cost me much, if it’s right it could be big. It reminds me to go with what I’ve built and know to be right. In fact you get so used to your own system you start to know when it’s likely to trigger, which I personally think is a very welcome sign because you are probably on your way to achieving the ultimate; being a rules-based discretionary trader.

For the record the same system is currently long crude oil at 91.78, long euro at 1.3261, short sterling at 1.5828, short yen at 1.2252.

It’s either wrong small and quickly, or it’s right large and slowly.

And what did it trigger Sunday night as the futures opened? Long S&P futures.

It may be right, it may be wrong, it really doesn’t matter either way to me, it’s just one trade within an overall winning system, but it came as no surprise just as many are calling for a top, that all I heard was that familiar ping! right on cue.

 

8 comments

Skip to comment form

  1. serendipitoustrader

    Jon – This was a brilliant post. Have you back tested the system? Do we know the percentage win, average dollar gain on win, average dollar loss when wrong,how many trades it generates per month?

    Also, what do you use as an exit strategy for this system?

    1. Jon Boorman

      That specific system been going 6 years, just under 400 trades, 46% win, using $100k portfolio value: avg win $3,700, avg loss $2,500, compound ann. return 13.0%, max drawdown 37%. Exit is invalidation of entry rationale (trend/trend strength, efficiency), or trailing stop reflecting risk/reward after given holding period.

  2. serendipitoustrader

    Thanks Jon. Curious here how you end up with a $2500 avg loss on a 100K portfolio. The way I want to trade is to risk about 1% (or less) of capital at risk per trade. I have a trading simulator spreadsheet and I can do various simulations using it. This system will work wonders if one can restrict the loss to 1% of capital or less.

    1. Jon Boorman

      Well I’m using very basic position sizing to come up with those figures for you, the actual position sizing strategy would be something personal to each traders objectives, what return you’re seeking, what you’re prepared to risk, risk of ruin etc. My example was assuming starting with $100k, it obviously would have grown over those six years, would currently be worth $210,000, and you will get slippage on your 1R risk. All of those figures were on basic assumption of just trading a single contract every single trade. Position sizing it properly could vary it enormously, but it just shows you a simple 1 contract every trade would still have been very effective.

  3. Ales

    Jon, I understand that you did not backtest your system, i.e., running the system’s rules through computer and historical data. However, you have been trading your system for 6 years with relative success. Have you been trading it live, or just on simulated account?

    What gives you confidence to trade it? Is it the past 6 years of data? Is it the fundamental trend-following math (avg. profit > avg. loss)? Is it your belief in your discrete trading skills?

    When you backtest several decades of data, you can be confident that the system held in the most various conditions in history. But 6 years is comparatively very little data and what has worked few years back, doesn’t have to work now. Where is the leap of faith comes from, in order to go live and keep trading regardless many losses one may encounter?

    1. Jon Boorman

      No, I did backtest the system. I haven’t always been able to trade it because of conflicts being employed as a broker and research analyst. The confidence to trade it comes from building it, you need to trade something that matches your personality and trading style. The 6 years is real-time data, the backtest would have been conducted on years prior to that, before it was implemented. any period that encompasses at least 1 or 2 bull and bear phases is normally sufficient.

  4. Gaurav S

    How do you backtest a system,.. is it through Trade navigator software

    1. Jon Boorman

      Yes, it’s what I’ve always used but there are many others available, you should research them all and find what will suit you best.

Leave a Reply